Re: [PATCH] f2fs: add fi->commit_lock to protect commit GCed pages

[Chao Yu <chao@kernel.org>]

On 2018/2/9 20:56, Yunlong Song wrote:
> As what I point in last mail, if the atomic file is not committed
> yet, gc_data_segment will register_inmem_page the GCed data pages.

We will skip GCing that page as below check:

- move_data_{page,block}
 - f2fs_is_atomic_file()
   skip out;

No?

Thanks,

> This will cause these data pages written twice, the first write
> happens in move_data_page->do_write_data_page, and the second
> write happens in later __commit_inmem_pages->do_write_data_page.
> 
> On 2018/2/9 20:44, Chao Yu wrote:
>> On 2018/2/8 11:11, Yunlong Song wrote:
>>> Then the GCed data pages are totally mixed with the inmem atomic pages,
>>
>> If we add dio_rwsem, GC flow is exclude with atomic write flow. There
>> will be not race case to mix GCed page into atomic pages.
>>
>> Or you mean:
>>
>>                     - gc_data_segment
>>                      - move_data_page
>>                       - f2fs_is_atomic_file
>> - f2fs_ioc_start_atomic_write
>>   - set_inode_flag(inode, FI_ATOMIC_FILE);
>>                       - f2fs_set_data_page_dirty
>>                        - register_inmem_page
>>
>> In this case, GCed page can be mixed into database transaction, but could
>> it cause any problem except break rule of isolation for transaction.
>>
>>> this will cause the atomic commit ops write the GCed data pages twice
>>> (the first write happens in GC).
>>>
>>> How about using the early two patches to separate the inmem data pages
>>> and GCed data pages, and use dio_rwsem instead of this patch to fix the
>>> dnode page problem (dnode page commited but data page are not committed
>>> for the GCed page)?
>>
>> Could we fix the race case first, based on that fixing, and then find the
>> place that we can improve?
>>
>>>
>>>
>>> On 2018/2/7 20:16, Chao Yu wrote:
>>>> On 2018/2/6 11:49, Yunlong Song wrote:
>>>>> This patch adds fi->commit_lock to avoid the case that GCed node pages
>>>>> are committed but GCed data pages are not committed. This can avoid the
>>>>> db file run into inconsistent state when sudden-power-off happens if
>>>>> data pages of atomic file is allowed to be GCed before.
>>>>
>>>> do_fsync:                GC:
>>>> - mutex_lock(&fi->commit_lock);
>>>>                      - lock_page()
>>>>                       - mutex_lock(&fi->commit_lock);
>>>>    - lock_page()
>>>>
>>>>
>>>> Well, please consider lock dependency & code complexity, IMO, reuse
>>>> fi->dio_rwsem[WRITE] will be enough as below:
>>>>
>>>> ---
>>>>    fs/f2fs/file.c | 3 +++
>>>>    fs/f2fs/gc.c   | 5 -----
>>>>    2 files changed, 3 insertions(+), 5 deletions(-)
>>>>
>>>> diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
>>>> index 672a542e5464..1bdc11feb8d0 100644
>>>> --- a/fs/f2fs/file.c
>>>> +++ b/fs/f2fs/file.c
>>>> @@ -1711,6 +1711,8 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp)
>>>>
>>>>        inode_lock(inode);
>>>>
>>>> +    down_write(&F2FS_I(inode)->dio_rwsem[WRITE]);
>>>> +
>>>>        if (f2fs_is_volatile_file(inode))
>>>>            goto err_out;
>>>>
>>>> @@ -1729,6 +1731,7 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp)
>>>>            ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 1, false);
>>>>        }
>>>>    err_out:
>>>> +    up_write(&F2FS_I(inode)->dio_rwsem[WRITE]);
>>>>        inode_unlock(inode);
>>>>        mnt_drop_write_file(filp);
>>>>        return ret;
>>>> diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
>>>> index b9d93fd532a9..e49416283563 100644
>>>> --- a/fs/f2fs/gc.c
>>>> +++ b/fs/f2fs/gc.c
>>>> @@ -622,9 +622,6 @@ static void move_data_block(struct inode *inode, block_t bidx,
>>>>        if (!check_valid_map(F2FS_I_SB(inode), segno, off))
>>>>            goto out;
>>>>
>>>> -    if (f2fs_is_atomic_file(inode))
>>>> -        goto out;
>>
>> Seems that we need this check.
>>
>>>> -
>>>>        if (f2fs_is_pinned_file(inode)) {
>>>>            f2fs_pin_file_control(inode, true);
>>>>            goto out;
>>>> @@ -729,8 +726,6 @@ static void move_data_page(struct inode *inode, block_t bidx, int gc_type,
>>>>        if (!check_valid_map(F2FS_I_SB(inode), segno, off))
>>>>            goto out;
>>>>
>>>> -    if (f2fs_is_atomic_file(inode))
>>>> -        goto out;
>>
>> Ditto.
>>
>> Thanks,
>>
>>>>        if (f2fs_is_pinned_file(inode)) {
>>>>            if (gc_type == FG_GC)
>>>>                f2fs_pin_file_control(inode, true);
>>>>
>>>
>>
>> .
>>
>